nrr, NATURAL HISTORY OF PLANTS. 



This family is also formed "by conuection," and such very nearly, in 

 the preceding account, as it was conceived by Blume,^ in 1847. He 

 united together, in fact: 1st. Sapindaceœ properly speaking, divided 

 into seven tribes," with flowers, some regular and others irregular ; 

 2nd. Dodoneacece ; 3rd. Accrincœ ; 4th. Hij^pocastaneœ ; 5th. Me- 

 Uosmeœ (Salnœ). In this group then are confounded the two 

 families which had, in 1789, been distinguished by A. L. de Jussieu ' 

 under the names of Sapindi and Accra, but to which unfortunately he 

 allied some Malpighiaccce as Tryallis, some Celasfraceœ as Hippocrateœ, 

 nnd some Ternstrœmiaceœ as Pclcca ; whilst Dodonœa was classed 

 among Terehinthacece , and Stapdujlca among the Rha?nni. In 1821, 

 KuNTH ■* divided the Sapindaceœ of equinoctial America into three 

 sections : Paidliniaceœ , Sapindaceœ proper and Bodoneaccœ, distinguish- 

 ing them from one another by the presence or absence of a scale 

 interior to the petals. De Candolle ^ preserved in 1824 this mode 

 of grouping which has absolutely no value, even for separating the 

 genera. Cambessedes,^ who resumed in 1829 the monographic study 

 of the Sapindaceœ., only divided them into Sapindeœ and Dodoneœ, in 

 whichhe was closely followed by Endlichee'' in 1838,andbyLiNDLET,^ 

 whose Sapindaceœ comprise all the groups admitted by Blume, except 

 Acereœ. The Staphjleœ, long included amongst Celastraccœ, were 

 finally separated by Bartling ^ and LiNDLEY,'"but as a distinct order, 

 and remained such until the epoch when Bentham and Hooker ^^ 

 made them only a tribe of Sapinducece. With them, this family con- 

 tains fi.ve sub-orders : Sapindeœ, Accrincœ, Bodoneœ , Melianthcœ , and 

 Staphijlcœ. The Meliantheœ had, before them,^- been ap^jroached, in 

 preference, to the Rutacece and Zijgophjllcœ. In this family, such as 

 they understand it, are united seventy-four genera, ^^ and they exclude, 

 as a distinct family,^* the Sahiaceœ, formerly classed in England ^^ 

 between Jfcnisjiermaceœ and Lardisabaleœ and previously placed by 



1 Eumphifi, iii. 91. » Old. Nat. 381. 



2 1. Sapindeic ; 2. Allophylleœ ; 3. Meli- '" Syiiops. (1829) 75 ; Veg. Kiiigd. 381, Ord. 

 cocceos ; 4. Cupanieas ; 5. Cossignieaj ; 6. Paul- 135. 



linieffi ; 7. ThouinieEe. " Om. (1862) 392, 409, Sutord. S. 



^ Gen. 246, 250, Ord. 5, 6. ''■' A. Juss. in Mem. Mtis. xii. 459. — Endl. 



< H. B. K. Gen. Nin: et Spec. v. 99. Gen. n. 6043.— I.indl. Ve,j. Elngd. 478. 



■' Priidr. i. 601. " Reduced for us to sixty-two. 



« Mém. Mus. xviii. (1829), i. " Gen. 413, Ord. 52. 



' Gen. 1066, Ord. 230. '* Hook. f. & Thoms. FI. Ind. i. 206. 



8 Veg. Kingd. (1846) 381, Ord. 135. 



